Non-small cell lung cancer (NSCLC) is the number one cancer killer of U.S. Veterans. Tobacco smoking is the major cause of NSCLC among Veterans. Meanwhile, tobacco smoking is also the primary cause of chronic obstructive pulmonary disease (COPD) that is a common benign disease in Veterans. Furthermore, Veterans with COPD are 4-6 times more likely to develop NSCLC than healthy individuals. Given the poor prognosis associated with advanced stage NSCLC, early detection of NSCLC in Veterans who are heavy smokers and have COPD will reduce the mortality from NSCLC. However, the current diagnostic techniques are either invasive or have poor accuracy. Here we propose to develop sputum-based non-coding RNA (ncRNA) biomarkers that can combine with the genomic probes for NSCLC early detection in heavy smokers and COPD patients among Veterans. Previously, we modified sputum induction protocol and established enrichment technique that can efficiently collect deep respiratory epithelial cells for molecular analysis of sputum. We also developed sputum-based genomic probes that can diagnose NSCLC with higher sensitivity compared with sputum cytology. However, the sensitivity of the probes is not high enough for clinical application. NcRNAs, particularly miRNAs, are emerging as potential biomarkers in cancer diagnosis. We recently demonstrated that ncRNAs are stably present in sputum and are potentially useful in diagnosis of lung cancer. Furthermore, we identified a set of ncRNA signatures including 21 miRNAs and five small nucleolar RNAs (snoRNAs) whose altered expressions are associated with early stage NSCLC. Because lung cancer is a heterogeneous disease and develops from a complex and multistep processes, the use of multiple types of biomarkers rather than only one class of biomarkers should have the potential to diagnose lung cancer with acceptable accuracy. Therefore, we hypothesize that analyzing the ncRNAs with the developed genomic probes in sputum will provide an accurate and noninvasive approach for early detection of NSCLC in Veterans. Three Aims are: 1, from the identified ncRNA signatures, optimizing a panel of highly specific and sensitive sputum biomarkers for early stage NSCLC, 2, determining the use of ncRNA biomarkers with genomic probes for NSCLC early detection in our existing sputum specimens, 3, validating the combined strategy for NSCLC early detection in an independent cohort. Upon completion, the work will lay a solid foundation for a clinical trial that will further validate the diagnostic value of the biomarkers. The use of the biomarkers that can identify NSCLC at its early stage will offer the best opportunity for effective treatments of the malignancy, and therefore will provide the best health care for Veterans diagnosed with lung cancer.